The Intergovernmental Panel on Climate Change (IPCC) concluded that climate change will likely increase forest fire severity, with fires that bur more intensely and spread more rapidly. NIH identified climate change and forest fires as a critical research need. According to the U.S. Forest Service, fire risk has already increased. Several studies suggest that forest fires affects health; however, the true health burden is unknown as no large- scale studies have been conducted. Current research is limited in scope and methods to estimate exposure to fire smoke. Scientific literature on ambient air pollutants indicates that some populations are more vulnerable than others to health responses. We hypothesize that some populations are more vulnerable to health impacts of forest fires under a changing climate. To estimate this vulnerability, new data and methods are needed. We propose to develop the necessary databases and methods so we can investigate the vulnerability to hospital admissions from forest fires in the western United States. In Aim 1 (exposure estimation), we will develop a new approach to integrate a chemical transport model, GEOS-Chem, with state-of-the-science forest fire emission inventories to estimate the contribution of forest fires to airborne particulate matter (PM). Our method generates daily, gridded estimates of PM2.5 from forest fire smoke and PM2.5 from other pollutant sources. In Aim 2 (risk estimation) we will develop a data set that links information on weather conditions, individual-level data on hospital admissions and co-morbidities, and community factors (e.g., unemployment) at the zip-code level with PM2.5 from forest fires. We will develop statistical models for estimating the association between exposure to forest fire smoke and risk of hospitalizations for persons >65 years. In this aim, we identify which populations are most vulnerable with respect to individual factors (age, race, sex, pre- existing conditions), region, community factors (e.g., unemployment) and environmental factors (e.g., weather). In Aim 3 (exposure and risk prediction through 2050), we will use the approach developed in Aim 1 and climate model projections to predict future area burned and forest fire emissions through 2050. These estimates will be used with chemical transport modeling to estimate future forest fire exposures. We will then use the approach developed in Aim 2 to predict the change in risk of hospitalizations, by vulnerability, under a changing climate accounting for the many sources of uncertainty in exposure and risk estimation. This study would advance understanding of how forest fires affect health in the present day and which populations are most vulnerable, and of how climate change could impact forest fires and thereby health and related vulnerabilities. Our findings would benefit efforts to manage forest fires, allow more targeted public health interventions, and aid effective allocation of resources, which is particularly important given the increased risk under climate change. The databases and methods developed would provide the research framework for studies of other environmental exposures, health outcomes, and climate change scenarios.